def plot_confusion_matrix(cm, labels_name, title):
cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis] # 归一化
plt.imshow(cm, interpolation='nearest', cmap="summer") # 在特定的窗口上显示图像
plt.title(title) # 图像标题
plt.colorbar()
num_local = np.array(range(len(labels_name)))
plt.xticks(num_local, labels_name) # 将标签印在x轴坐标上
plt.yticks(num_local, labels_name, rotation=90) # 将标签印在y轴坐标上
plt.ylabel('True label')
plt.xlabel('Predicted label')
predicted = np.argmax(resultall, 1)
confu_matrix = cal_confu_matrix(np.array(predicted),
np.array(lableforconfusion),
class_num=2)
print(confu_matrix)
labels_name = ["inactive CNV", "active CNV"]
plot_confusion_matrix(confu_matrix, labels_name, "Confusion Matrix")
plt.savefig('./savedimg/confusionmatrix_octa.png',
dpi=300,
bbox_inches='tight')
plt.show()
metrics(confu_matrix, save_path="./savedimg/")
for i in range(n_classes):
fpr[i], tpr[i], _ = roc_curve(y_test[:, i], y_score[:, i])
roc_auc[i] = auc(fpr[i], tpr[i])
label='Doctor3',
tick_label=name_list,
fc='plum')
plt.legend(loc="upper left", bbox_to_anchor=(0.0, 1.1), ncol=4)
plt.savefig('./savedimg/OCT/CNN_Doctor_precission_comparision_EACH.png',
dpi=300)
plt.show()
#draw recall bar pic each
from metrics import calc_metrics, cal_confu_matrix, metrics
import numpy as np
aaaaaa = np.array(data['A'])
confu_matrix = cal_confu_matrix(np.array(data['A']),
np.array(data['D']),
class_num=4)
print(confu_matrix)
metrics(confu_matrix, save_path="./savedimg/OCT/doctor1")
confu_matrix = cal_confu_matrix(np.array(data['B']),
np.array(data['D']),
class_num=4)
print(confu_matrix)
metrics(confu_matrix, save_path="./savedimg/OCT/doctor2")
confu_matrix = cal_confu_matrix(np.array(data['C']),
np.array(data['D']),
class_num=4)
print(confu_matrix)
metrics(confu_matrix, save_path="./savedimg/OCT/doctor3")